Prostate cancer remains the most commonly diagnosed malignancy in men in the United States and a leading cause of cancer related mortality. Nomograms and multi-variable tables identify men at high risk for recurrence following definitive local therapy but provide little insight as to the specific biology driving aggressive disease. Recent findings have highlighted the role for genetic translocations in lethal disease and molecular-based prognostic and pathway signatures have the potential to guide multi-modality therapy and improve the potential for cure. Our proposal will validate established RNA-, DNA-, and microRNA prognostic and pathway signatures in both a retrospective and prospective cohort of men with high-risk prostate cancer. We will first comprehensively assay a retrospective cohort of 200 men who have had prostatectomy, have frozen and paraffin embedded tissue available, and have mature clinical follow up. This cohort will be processed through a centralized system including the CALGB pathology coordinating office, assayed in laboratories with established expertise in comprehensive RNA-, DNA-, and microRNA analysis, and results will be analyzed in the CALGB Statistical Center. This cohort will provide the initial validation of prognostic and predictive signatures and determine how these signatures relate to ETS-family member and RAF translocations. The most informative signatures will be re-derived by the statistical center, re-applied to the retrospective dataset to confirm reproducibility of the computational models, and then applied to a samples from CALGB 90203, a prospective, randomized phase III trial that will randomize 750 men between immediate surgery or neoadjuvant docetaxel and androgen deprivation therapy followed by surgery. In addition, computational investigators will perform integrative analysis of RNA-, DNA-, and microRNA data so as to distinguish "driver" pathways responsible for the lethal phenotype from "bystander" pathways that are correlated with aggressiveness but not causative. Finally, promising, validated signatures will be adapted to paraffin-embedded tissue so as to develop clinically deployable bioassays. Our proposal leverages two outstanding cohorts, the scientific and computational expertise in five independent laboratories, the established capabilities of industrial partners, and the statistical rigor of the CALGB to validate prognostic and pathway signatures in high-risk prostate cancer. All data will be shared through GEO and other emerging databases so as to ensure that along with the insight resulting from our own analysis, our data will be an outstanding resource for the larger investigative community interested in the validation of prognostic signatures in high-risk prostate cancer.